Fluid Mechanics, SG2214, HT2010

September 14, 2010

Exercise 5: Exact Solutions to the Navier-Stokes Equations I

Example 1: Plane Couette Flow

Consider the flow of a viscous Newtonian fluid between two parallel plates located at y = 0 and y = h. The
upper plane is moving with velocity U . Calculate the flow field.

Assume the following:

Steady flow:
∂
=0
∂t
Parallel, fully-developed flow:
∂ui
v = 0, =0
∂x
Two-dimensional flow:
∂
w = 0, =0
∂z
No pressure gradient:
∂p
=0
∂xi
The streamwise Navier-Stokes equation is
∂u 1
+ (ū · ∇)u = − ∇p + ν∇2 u ,
∂t ρ
can be simplified using the above assumptions. We get

∂2u ∂u
=0 ⇒ =A ⇒ u = Ay + B .
∂y 2 ∂y
With the boundary conditions

u(0) = 0 ⇒ B = 0, u(h) = U ⇒ A = U/h

we finally obtain
Uy
u(y) = .
h

Example 2: Plane Poiseuille Flow (Channel Flow)

Consider the flow of a viscous Newtonian fluid between two solid boundaries at y = ±h driven by a constant
pressure gradient ∇p = [−P, 0, 0]. Show that
P 2
u= (h − y 2 ), v = w = 0.
2µ

Navier-Stokes equations: 
 ∂ ū + (ū · ∇)ū = − 1 ∇p + ν∇2 ū
∂t ρ

∇ · ū = 0.
Boundary conditions:
ū(y = ±h) = 0

1
 Y

Figure 1: Coordinate system for plane Poiseuille flow.

∂ ū
• We are considering stationary flow and thus = 0.
∂t
• The constant pressure gradient implies ū = ū(y). Changes of ū in x, z would require a changing pressure
gradient in x, z.
∂v
• The continuity equation ∇ · ū reduces to = 0. The boundary condition v(y = ±h) = 0 then implies
∂y
v = 0.

Consider the spanwise (z) component of the Navier-Stokes equations:

∂w ∂2w
v =ν 2 ⇒ w = c1 y + c2
|{z} ∂y ∂y
=0

The boundary conditions w(y = −h) = w(y = h) = 0 imply c1 = c2 = 0 and thus w = 0. We can conclude
that ū = [u(y), 0, 0].

Consider now the streamwise (x) component of the Navier-Stokes equations:

P ∂2u ∂u P P 2
0= +ν 2 ⇒ = − y + d1 {µ = ρν} ⇒ u(y) = − y + d1 y + d2
ρ ∂y ∂y νρ 2µ
The boundary conditions at y = ±h give
P 2 P 2
0=− h + d1 h + d2 and 0 = − h − d1 h + d2
2µ 2µ
P 2
We can directly conclude that d1 = 0 and this gives d2 = h . The solution is thus
2µ

P 2
u= (h − y 2 ), v=w=0.
2µ

Energy Dissipation in Poiseuille Flow

a) Calculate the dissipation function for the plane Poiseuille flow computed above,

P 2
u= (h − y 2 ), v = w = 0,
2µ
or in terms of the bulk velocity U
3U 2
u= (h − y 2 ), v = w = 0.
2h2

2
 The mass-flow rate through the channel is
Z h
Q= udy = 2U h .
−h

The dissipation function is defined as (dissipation to heat due to viscous stresses)

∂ui
Φ = τij .
∂xj

For incompressible flows, it can be re-written as

∂ui
Φ = τij = 2µeij (eij + ξij ) = 2µeij eij ,
∂xj

where we used the fact that eij ξij = 0.

 ∂u 
0 ∂y 0
The deformation tensor for the Poiseuille flow becomes eij = 1/2  ∂u
∂y 0 0 and therefore
0 0 0
" 2  2 #  2
1 ∂u 1 ∂u ∂u
Φ = 2µ + =µ .
2 ∂y 2 ∂y ∂y

b) Calculate the total dissipation for unit area

Z h Z h 2
3U 6µU 2
φ= Φ dy = µ − 2 y dy = .
−h −h h h

c) Write the mechanical energy equation for this flow. Integrate over the channel width and relate the
total dissipation φ to the pressure gradient and the mass flux.
The mechanical energy equation is obtained by multiplying the Navier-Stokes equations by ui (the
energy is ρ(1/2)ui ui ). One gets
 
D 1 ∂p ∂τij
ρ ui ui = ρFi ui − ui + ui .
Dt 2 ∂xi ∂xj

Considering the Poiseuille flow and re-writing the last term as

∂τij ∂ui τij
ui = − Φ,
∂xj ∂xj

the energy equation reduces to

∂
0 = uP + (uτxy ) − Φ.
∂y
Integrating across the channel each term in the expression above, one obtains for the first term
Z h Z h
uP dy = P udy = QP,
−h −h

where Q is the flow rate. This term represents the work rate by pressure forces.
The second term Z h
∂ h
(uτxy ) dy = [(uτxy )]−h = 0
−h ∂y
due to the no-slip boundary conditions.

3
 Rh
The third term is the total dissipation φ = −h
Φ dy defined above. Summarising
Z h
0 = QP − Φ dy.
−h

One can check the results, using the expression for φ obtained in b). Just recall that
Z h
Q= udy = 2U h,
−h

3µU
and the pressure gradient can be expressed in terms of U as P = h2 . Therefore QP = 6µU 2 /h = φ.

Example 3: Poiseuille Flow (Pipe Flow)

Consider the viscous flow of a fluid through a pipe with a circular cross-section given by r = a under the
∂p
constant pressure gradient P = − . Show that
∂z
P 2
uz = (a − r2 ) ur = uθ = 0.
4µ

Figure 2: Coordinate system for Poiseuille flow.

Use the Navier-Stokes equations in cylindrical coordinates (see lecture notes)

 
∂ur u2 1 ∂p ur 2 ∂uθ
+ (ū · ∇)ur − θ = − + ν ∇2 ur − 2 − 2
∂t r ρ ∂r r r ∂θ
 
∂uθ ur uθ 1 ∂p 2 2 ∂ur uθ
+ (ū · ∇)uθ + =− + ν ∇ uθ + 2 − 2
∂t r ρ r ∂θ r ∂θ r
∂uz 1 ∂p
+ (ū · ∇)uz = − + ν∇2 uz
∂t ρ ∂z
1 ∂ 1 ∂uθ ∂uz
(r ur ) + + =0.
r ∂r r ∂θ ∂z
∂p ∂p
We know that = 0 and = 0 and can directly see that ur = uθ = 0 satisfys the two first equations.
∂θ ∂r
From the continuity equation we get
∂uz
=0 ⇒ uz = uz (r, θ) only.
∂z
Considering a steady flow we get from the axial component of the Navier-Stokes equations
∂uz ∂uz 1
(ū · ∇)uz = uz ⇒ uz = P + ν∇2 uz .
∂z ∂z ρ

4
 ∂uz
But we know that = 0 from the continuity equation. We get
∂z
1 ∂ ∂uz P ∂ ∂uz P
∇2 uz = (r )=− ⇒ (r )=− r
r ∂r ∂r µ ∂r ∂r µ
Integrate once in r gives
∂uz P 2 ∂uz P c1
r =− r + c1 ⇒ =− r+ ,
∂r 2µ ∂r 2µ r
and integrating again we get
P 2
uz = − r + c1 ln(r) + c2 using the boundary conditions uz (r = 0) < ∞ ⇒ c1 = 0 .
4µ

P a2
We also have uz = 0 at r = a and this gives c2 = and we finally get
4µ

P 2
uz = (a − r2 ) .
4µ

Example 4: Asymptotic Suction Boundary Layer

Calculate the asymptotic suction boundary layer, where the boundary layer over a flat plate is kept parallel
by a steady suction V0 through the plate.

Assumptions:
Two-dimensional flow:
∂
= 0, w=0
∂z
Parallel, fully-developed flow:
∂
=0
∂x
Steady flow:
∂
=0
∂t
Momentum equations:  2 
∂u ∂u ∂u 1 ∂p ∂ u ∂2u
+u +v =− +ν + 2
∂t ∂x ∂y ρ ∂x ∂x2 ∂y
 2 
∂v ∂v ∂v 1 ∂p ∂ v ∂2v
+u +v =− +ν + 2
∂t ∂x ∂y ρ ∂y ∂x2 ∂y
Normal momentum equation gives
∂p
=0
∂y
Boundary conditions:
y=0: u = 0, v = −V0
y→∞: u → U∞
Continuity gives
∂u ∂v
+ =0 ⇒ v = −V0
∂x ∂y
Streamwise momentum equation at y → ∞

∂U∞ 1 ∂p ∂ 2 U∞
−V0 =− +ν
∂y ρ ∂x ∂y 2

5
 ∂p
⇒ =0
∂x
Resulting streamwise momentum equation

∂u ∂2u ∂2u V ∂u
−V0 =ν 2 ⇒ 2
=− 0
∂y ∂y ∂y ν ∂y
Characteristic equation
V0 V0
λ2 = − λ ⇒ λ1 = 0, λ2 = −
ν ν
u(y) = A + Be−V0 y/ν
With the boundary conditions at y = 0 and y = ∞ we get
 
u(y) = U∞ 1 − e−V0 y/ν .

Example 5: Flow on an Inclined Plate

Two incompressible viscous fluids flow one on top of the other down an inclined plate at an angle α (see figure
3). They both have the same density ρ, but different viscosities µ1 and µ2 . The lower fluid has depth h1 and
the upper h2 . Assuming that viscous forces from the surrounding air is negligible and that the pressure on
the free surface is constant, show that
 
1 g sin(α)
u1 (y) = (h1 + h2 )y − y 2 .
2 ν1

11111
00000
h2 111
00000
11111000 0000000 Y
1111111
000000000000000000000000000
111111111111111111111111111
000
111
00000
11111 0000000
1111111
000000000000000000000000000
111111111111111111111111111
0000000
1111111
00000
11111 000000000000000000000000000
111111111111111111111111111
0000000000000000000000000000
1111111111111111111111111111
11
00 0000000
1111111
h11111
00000 000000000000000000000000000
111111111111111111111111111
0000000000000000000000000000
1111111111111111111111111111
0000000
1111111
000000000000000000000000000
111111111111111111111111111
00000
11111
1
0000000000000000000000000000
1111111111111111111111111111
00000
11111 0000000
1111111
000000000000000000000000000
111111111111111111111111111
111
000 0000000000000000000000000000
1111111111111111111111111111
0000000
1111111
000000000000000000000000000
111111111111111111111111111
0000000000000000000000000000
1111111111111111111111111111
0000000
1111111
000000000000000000000000000
111111111111111111111111111
0000000000000000000000000000
1111111111111111111111111111
0000000
1111111
000000000000000000000000000
111111111111111111111111111
0000000000000000000000000000
1111111111111111111111111111
0000000
1111111
000000000000000000000000000
111111111111111111111111111
0000000000000000000000000000
1111111111111111111111111111
0000000
1111111
000000000000000000000000000
111111111111111111111111111
0000000000000000000000000000
1111111111111111111111111111
0000000
1111111
000000000000000000000000000
111111111111111111111111111
0000000000000000000000000000
1111111111111111111111111111
0000000
1111111
000000000000000000000000000
111111111111111111111111111
0000000000000000000000000000
1111111111111111111111111111
000000000000000000000000000
111111111111111111111111111
0000000000000000000000000000
1111111111111111111111111111
000000000000000000000000000
111111111111111111111111111
0000000000000000000000000000
1111111111111111111111111111
000000000000000000000000000
111111111111111111111111111
0000000000000000000000000000
1111111111111111111111111111
000000000000000000000000000
111111111111111111111111111
0000000000000000000000000000
1111111111111111111111111111
0000000000000000000000000000
1111111111111111111111111111
0000000000000000000000000000
1111111111111111111111111111
X

Figure 3: Coordinate system for flow down an inclined plate.

Make the ansatz ū1 = [u1 (y), 0, 0] and ū2 = [u2 (y), 0, 0]. The continuity equation

∂u ∂v ∂v
+ =0 gives = 0 ⇒ v = c and the boundary condition at y = 0 give v = 0.
∂x ∂y ∂y
• Layer 1:
1 ∂p1
N–S · ēy : 0 = − − g cos(α) ⇒ p1 = −ρg cos(α)y + f1 (x)
ρ ∂y
1 ′ d2 u 1 ′
N–S · ēx : 0 = − f1 (x) + ν1 2 + g sin(α) ⇒ f1 (x) = c1
ρ dy
• Layer 2:
1 ∂p2
N–S · ēy : 0 = − − g cos(α) ⇒ p2 = −ρg cos(α)y + f2 (x)
ρ ∂y
1 ′ d2 u 2 ′
N–S · ēx : 0 = − f2 (x) + ν2 2 + g sin(α) ⇒ f2 (x) = c2
ρ dy

6
The pressure at the free surface y = h1 + h2 is p0 :
′
p0 = −ρg cos(α)(h1 + h2 ) + f2 (x) ⇒ f2 = p0 + ρg(h1 + h2 ) cos(α) ⇒ f2 = 0
The pressure is continuous at y = h1 :
′
p0 + ρgh2 cos(α) = −ρgh1 cos(α) + f1 (x) ⇒ f1 = p0 + ρg(h1 + h2 ) cos(α) ⇒ f1 = 0
This gives the pressure:
p1 (y) = p2 (y) = p(y) = −ρ g cos(α)y + p0 + ρ g cos(α)(h1 + h2 )
We now have two momentum equations in x:
d2 u 1
0 = ν1 + g sin(α) (1)
dy 2
d2 u 2
0 = ν2 + g sin(α) (2)
dy 2
And four boundary conditions:
BC1: No slip on the plate: u1 (0) = 0

du2 
BC2: No viscous forces on the free surface: µ2 =0
dy y=h1 +h2
 
du1  du2 
BC3: Force balance at the fluid interface: µ1 = µ2
dy y=h1 dy y=h1
BC4: Continous velocity at the interface: u1 |y=h1 = u2 |y=h1
du1 g g 2
(1) ⇒ = − y sin(α) + c11 ⇒ u1 = − y sin(α) + c11 y + c12
dy ν1 2 ν1
du2 g g 2
(2) ⇒ = − y sin(α) + c21 ⇒ u2 = − y sin(α) + c21 y + c22
dy ν2 2 ν2
BC1 ⇒ c12 = 0
g g
BC2 ⇒ µ2 (− (h1 + h2 ) sin(α) + c21 ) = 0 ⇒ c21 = (h1 + h2 ) sin(α)
ν2 ν2
g g µ2 g
BC3 ⇒ µ1 (− y sin(α)+c11 ) = µ2 (− y sin(α)+c21 ) {µ = νρ} ⇒ c11 = c21 = (h1 +h2 ) sin(α)
ν1 ν2 µ1 ν1
g 2 g g 2 g
BC4 ⇒ − h sin(α) + (h1 + h2 ) sin(α)h1 = − h sin(α) + (h1 + h2 ) sin(α)h1 + c22
2 ν1 1 ν1 2 ν2 1 ν2
 2  
h 1 1
⇒ c22 = g sin(α) 1 − (h1 + h2 )h1 −
2 ν2 ν1
This gives us the velocities:
g 2 g
u1 (y) = − y sin(α) + (h1 + h2 ) sin(α)y
2 ν1 ν1
 
g sin(α) 1
u1 (y) = (h1 + h2 )y − y 2
ν1 2
 2  
g sin(α) 2 g sin(α) h1 1 1
u2 (y) = − y + (h1 + h2 )y + g sin(α) − (h1 + h2 )h1 −
2 ν2 ν2 2 ν2 ν1
   2  
g sin(α) 1 h 1 1
u2 (y) = (h1 + h2 )y − y 2 + g sin(α) 1 − (h1 + h2 )h1 −
ν2 2 2 ν2 ν1
The velocity in layer 1 does depend on h2 but not on the viscosity in layer 2. This is because the depth
is important for the tangential stress boundary condition at the interface, unlike the viscosity. There is no
acceleration of the upper layer and thus the tangential stress must be equal to the gravitational force on the
upper layer which depends on h2 but not on ν2 .